Long duration of the ~56 Myr Sturtian Snowball Earth event suggests missing link in geologic carbon cycle.

There were at least two major glaciation events (i.e., “Snowball Earth” events) in the Neoproterozoic Era when ice sheets dominated Earth’s surface. Radioisotopic dating indicates the first, Sturtian glaciation lasted ~56 Myr, and the second, Marinoan glaciation lasted ~4 Myr.Why do the two glaciation events – which are only separated by a ~22 Myr interglacial gap – have a ~14x difference in duration? To first order, the glacial termination, and thus duration, is determined by (1) changes in the albedo of Earth’s surface and/or (2) changes in greenhouse warming from the atmosphere, likely driven by enhanced CO2 and the geologic carbon cycle.Here, we simulated the evolution of atmospheric CO2 via the geologic carbon cycle during the Sturtian and Marinoan glaciation events to determine what conditions could explain the difference in their duration. While the 4 Myr Marinoan glaciation was reproduced in >30% of model runs with a variety of model parameter values, we find that only 0.05% of model runs reproduced the 56 Myr Sturtian glaciation. The Sturtian model runs require very low levels of CO2 outgassing from volcanos and extremely efficient seafloor weathering (which consumes CO2) to keep atmospheric CO2 levels low enough to sustain glacial conditions for 56 Myr. To reproduce the Marinoan glaciation, the opposite is required: a 1.6x increase in CO2 outgassing and a 10x decrease in seafloor weathering.What could cause such drastic changes in CO2 outgassing and seafloor weathering in successive glaciations separated by only 22 Myr? Possible explanations relate to the Franklin Large Igneous Province (LIP), the depth of mid-ocean ridges, and high-temperature anhydrite production in the seafloor; however, none of these explanations are directly indicated by geologic evidence.Therefore, the differing durations of the two glaciation events – and particularly the long duration of the 56 Myr Sturtian – indicate that an important aspect of the Neoproterozoic carbon cycle is not being captured. We suggest several possibilities and look forward to open discussions that may illuminate the solution.